Transactional Log Analysis for OPAC redesign at CU Boulder #RMRIUG
Matt Hamilton | June 12, 2009This was the morning presentation from the Rocky Mountain Regional Innovative Users Group Summer Workshop. Presented June 12, 2009 at the University of Colorado at Denver Health Sciences Library.
Jennifer Knievel and Jina Wakimota – University of Colorado at Boulder Chinook OPAC redesign
data to examine impact of redesign – took pulse before and after
old interface cluttered and forced user to choose index first
new interface clean, “Google experience” with access to more advanced features for researcher needs
Not just presentation — back end work, re-presentation of metadata
research q: would keyword default increase keyword searching?
are users using catalog less and migrating to Google more?
abandoned using 225 or 229 fields for periodical search– changed to Scope instead so that periodical-specific search could be enabled
reindexed database (had not been done in 10 years) — had been changes in Marc, Millenium features, etc.
Methods
Transactional Log analysis – unbiased, unobstrusive, lends itself to longitudinal analysis
Text Searches- Author, Title, Keyword, Subject
slight increase in keyword searches (about 4%)
slight decrease in Author searches (about 4%)
everything else stayed about the same with a spike in Title searches just before redesign (coincided with intro of periodical scope- search for journal titles)
Did a T-test for statistical significance
beginning and end of semester more known search, mid-semester more discovery (more keyword)
subject lowest used – no surprise
author low use
title and keyword rising slowly over time
downside of transactional analysis– can’t assess user success
Number Searches – Call #, ISSN/ISBN only
dramatic increase in LC call number searches– highest in the beginning of the semester (locating the items given to find perhaps?)
increase in ISSN searching before WebBridge implementation – correlated with making it easier
Total Searches
Catalog use steady – did not see decrease in usage, saw a spike in usage when new OPAC first implemented but then it went back to previous numbers






